737NGs have cracked 'pickle forks' after finding several in the jets.

[Aihkio]

At their present level computers still cannot design anything even close to the complexity level of an airliner. On the other hand without computer analysis the now used structures could not be designed.

[UltraFan]

Does Aeroflot have any affected 737s? I just don't know what to look for on all those "fleets" websites. Could anyone help, please?

[Momoe]

Bueno Hombre,

The airframe isn't ancient, the design is which is the problem. Boeing stuck with the 737 due to it's popularity and to avoid a full re-certification associated with a new design.

As the well worn Irish saying goes, 'If you're going there, I wouldn't start from here',

[Chu Chu]

Originally Posted by Aihkio

At their present level computers still cannot design anything even close to the complexity level of an airliner.

Much less eliminate the possibility of unanticipated fatigue cracks, which may relate the the interaction of design and fabrication techniques.

Of course, larger engines and increased weights on an old-design airframe can expose new issues. But if anything, that seems less likely to happen than with a wholly new design (assuming the appropriate analysis is done when the upgrades are made, of course). There are lots of possible advantages to starting with a clean slate, but this doesn't seem like one of them.

[Pilot DAR]

Though I am not expert in B 737 structure, I know of other types which have had vulnerable fastener holes reworked in place to prevent anticipated cracking based upon experience with that type. Twin Otter wing strut attach structure in the wing is an example. In some cases, a crack or other defect can be allowed to continue in service with inspection. I have approved two aircraft this way. Analysis of the defect showed that the load could be carried in the structure anyway, and inspection for further crack growth was easy. I'm working on such an approval right now on a smaller GA type. In the worst case, allowing the defect to remain in service unrepaired could involve a small gross weight, or maximum landing weight penalty. Part replacement is a great idea, and the best solution, if it is practical, and the replacement parts are available. If not, more lateral thinking may be appropriate.

I was a part of an inspection team for a prepurchase inspection of an Air New Zealand DC-8 30 years ago, which had a crack in a landing gear trunnion. The crack was monitored using a Douglas approved inspection program, and the plane remained in service. Air New Zealand had the replacement part in stock, but it would be a huge task to install it.

I have no idea what Boeing will propose for this defect, and in the present climate, Boeing will have to be very robust in their proposal. But I'm not imagining a wholesale "replace every pickle fork out there" program.

[Dave Therhino]

The type of activity you described led 35 years ago to formally issued FAA policy that prohibits approval of revenue "flight with cracks" for primary structure or fail safe features, although some inspectors or offices may not consistently follow the formal policy. The AD program for the pickle fork fitting undoubtedly will continue to involve monitoring for cracks beginning at some threshold, with repetitive inspections, with the intent of providing multiple opportunities to detect the cracks before they can go critical. However, once any cracking is found an appropriate repair or replacement will be required. In this case nobody has yet identified a practical and effective repair scheme, so replacement for now is the corrective action once a crack is detected. Since the NG is out of production and the cracks seem to appear later in the life of the airplane, Boeing may just stick with the existing design and live with the potential for one replacement required in the life of an airplane. I have not heard anything about development of a redesigned fitting, but it's possible one is in work.

I'm hearing rumors that cracks have been found in adjacent locations on a few airplanes ...

[ozaub]

Courtesy of Grandfather, 737 is only significant airliner still in service that has not been subject to a proper full scale fatigue test. Fail-safe fatigue integrity was originally certificated on basis of similarity with 707 and 727. There was also a fatigue test of rear fuselage to prove integrity of rear pressure bulkhead, which was misapplied to prove damage tolerance of fuselage lap splices. All of which ended in tears (pun!!). See LESSONS FROM ALOHA or http://www.iasa.com.au/folders/Safet...fromaloha.html

Aloha story mentions Boeing’s untimely letter immediately before accident - saying nothing could go wrong! I’ll share it with anyone interested; just PM me. No, I didn’t steal from Australian Government files. Got mine from Boeing’s file copy tabled at NTSB inquiry. Letter is dated two weeks before accident, so in days of snail mail its arrival was even closer.

Apparently 737 NG had a better fatigue test of fuselage. See https://www.newsweek.com/boeings-737...problems-63629 including:

“The Aviation Safety Institute's Pat Duggins told me that 38 changes were made to the fuselage before the NG went into production. And then, to make sure that no critical weaknesses remained, Boeing took a standard 737NG fuselage off the 737 assembly line at Wichita, Kans., and tested it to the breaking point. The airframe was pushed through the equivalent of 225,000 cycles (three times the assumed safe life of 75,000 cycles for the NG series) on short duration flights—exactly the way Southwest, for example, uses the 737.The problem, Duggins says, was that the test fuselage did not represent the realities of everyday flight. It lacked a wing box, the core load-bearing part of the wings where they meet the fuselage, and also the landing gear, which transmits particularly forceful stresses to the fuselage on every landing. In addition, he does not believe that the design changes would have given the NG series fuselage a significantly longer life. He raised doubts, when I talked to him, that the tests met Boeing's design requirements. (Boeing disagrees. It asserts that the testing "provides a realistic simulation of complete flights," and adds that "all these loads were represented"; Boeing did not specifically answer my questions about whether the wing box and landing gear were part of the airframe that was tested.)”

I doubt efficacy of this test to replicate fatigue stresses in pickle forks. Furthermore Boeing probably thought fatigue of forks unlikely because they are primarily in compression.

Main job of the forks is to transfer weight of fuselage to wing, putting forks in compression. Secondary tension stresses come from fuselage pressurisation ie fuselage lifting away from wing. So, tension in forks arises from pressurisation together with negative G, and is probably not well replicated in fatigue test spectrum.

With forks we may have dodged a bullet, much as we did with Section 41 cracking on 747. See https://www.flightglobal.com/FlightP...20-%200317.PDF Just by luck widespread cracks were first found in wreckage of JAL 123 in 1985, though nothing to do with crash. Likewise we almost dodged a bullet with 737 lap joints; except for poor Clarabelle Lansing

All of this I had in mind when penning https://www.smh.com.au/national/self...03-p536wk.html Plus I was irked by CASA and Qantas disingenuous platitudes downplaying significance of pickle fork cracking.

[pilot9248]

Thanks ozaub for the informative post. At first, I thought it excessive to demand inspections be carried out on all 737NGs regardless of flight cycles. However, with the latest developments of cracking apparently having been found in other locations than the inspection areas of the AD-mandates MOM, in conjunction with findings on aircraft below 22k flight cycles, I am becoming increasingly doubtful of Boeing’s approach. The multitude of existing issues in the area of Body Station 663 on all 737 minor models up to and including the NG does not exactly dispel my doubts. It is a very busy area in terms of loads going through the airframe, so one would hope that Boeing are being sufficiently conservative in their analysis. I believe it is safe to say that if an NG should ever suffer an accident due to failure of the joints in question, that would spell the end for Boeing (and the FAA as we know it). Hopefully that will never happen, but as engineers we shouldn’t have to summon hope.

[Just a spotter]

According to The Irish Times, 5th Nov 2019, Ryanair has confirmed 3 aircraft are affected by the pickle fork cracks;

The grounded planes have the registration numbers EI-DCL, EI-DAL and EI-DCJ. All three are more than 15 years old.

https://www.irishtimes.com/business/...lage-1.4074196

JAS

[ManaAdaSystem]

Originally Posted by MickG0105

Would you happen to know what the regos are for those two SAS aircraft?

Sorry, took a while to get hold of my SAS man.
LN-RPK. 737-700 with winglets. Has been with SAS since new. 19 years old.
SE-RET. 737-700 without winglets. Joined SAS in 2012. Previously with Virgin Blue/Virgin Australia reg VH-VBM. 17 years old.

[ozaub]

Thanks Turbo. I’m not advocating inspection before further flight; just inspect all 737s, oldest first.
It’s ludicrous for CASA and Qantas to insist there’s no risk below an arbitrary threshold. We simply don’t know; because 737 is only significant airliner that's never been properly fatigue tested. Except with passengers on board.
According to AD it only costs $170 to check for cracks. Just do it, instead of prattling about safety being paramount.

[MickG0105]

Originally Posted by ozaub

MOM gave airlines four weeks advance notice of the AD. Ample time to inspect most aircraft and have a "we've done it already" PR coup when AD came out.

That's not correct. The MOM was issued on 30 Sep 2019 1731 US PACIFIC TIME / 01 Oct 2019 0031 GMT. The AD was issued October 3, 2019. That's at best four days, not four weeks.

[MickG0105]

Originally Posted by ManaAdaSystem

Sorry, took a while to get hold of my SAS man.
LN-RPK. 737-700 with winglets. Has been with SAS since new. 19 years old.
SE-RET. 737-700 without winglets. Joined SAS in 2012. Previously with Virgin Blue/Virgin Australia reg VH-VBM. 17 years old.

Thank you for the follow up.

So it looks we're seeing the cracking problem manifest itself in some aircraft built before 2004 but right across the whole NG range (-600, -700, -800 and -900), and both with winglets and without.

This may simply be a coincidence but Boeing moved the Commercial Airplanes - Fabrication Division's complex machining from their Auburn, Washington plant to their Portland, Oregon facility between April 2003 and January 2005. Among the parts packages that were transitioned from Auburn to Portland were pickle forks. The move gets mentioned in a Boeing Frontiers article. It's probably unrelated but you never know.

[megan]

It is a very busy area in terms of loads going through the airframe, so one would hope that Boeing are being sufficiently conservative in their analysis

90,000 cycles is being quoted as the expected design life of the forks, fell some ways short it seems if that were the case.

[pilot9248]

Mick, I believe ozaub is referring to the previous MOM that informed operators of the issue, not the MOM that contains the inspection instructions.
I also do not advocate inspecting all aircraft before further flight. Generally, if it involves fatigue cracking, inspecting “too early” may be counterproductive.
Having said that, fatigue analysis is notoriously error-prone. I have seen exercises conducted throughout the industry to benchmark predictive analysis with subsequent fatigue testing to validate the results, and some of the results are so far off (to the wrong side) it’s truly shocking. That’s one of the reasons why testing is required for certification, even though the OEMs have been trying to get away with less testing and more analysis and modelling. Stating the obvious, testing is expensive.

[DaveReidUK]

Originally Posted by MickG0105

That's not correct. The MOM was issued on 30 Sep 2019 1731 US PACIFIC TIME / 01 Oct 2019 0031 GMT. The AD was issued October 3, 2019. That's at best four days, not four weeks.

The issue date of the AD was September 30. October 3 was the "effective from" date.

[Australopithecus]

This issue makes me wonder about inspection scheduling to airline convenience. Since the FAA allows no revenue flights with cracks in primary structures, how long before an avaricious lawyer sues an airline for operating aircraft that were not inspected just prior to their last revenue flight after which cracks were then discovered? I am told by engineers that my airline is avoiding inspections at airports it would be difficult and costly to do a repair for example.

That an airline has the option to inspect, knowing that some of their fleet have cracks, but does not for convenience smacks of reckless endangerment. (I am not going to qualify or attempt to quantify any risk. But I do note that cracks result in an immediate grounding, which is all a jury needs to hear.)

[MickG0105]

Originally Posted by theturbofantastic

Mick, I believe ozaub is referring to the previous MOM that informed operators of the issue, not the MOM that contains the inspection instructions.

Unless I'm mistaken the previous MOM first notifying operators of the issue - MOM-MOM-19-0530-01B(R1) - was released on 27 September 2019.

Originally Posted by DaveReidUK

The issue date of the AD was September 30. October 3 was the "effective from" date.

Yes, quite correct. My mistake.

So, the upshot of all that is that there was about three days between the first Boeing MOM and the AD, meaning that

Originally Posted by ozaub

MOM gave airlines four weeks advance notice of the AD. Ample time to inspect most aircraft and have a "we've done it already" PR coup when AD came out.

​
is not correct.

[artee]

Apparently Lion Air found cracks on two 737s with less than 22,000 flights.
https://www.smh.com.au/business/comp...06-p53828.html

[ozaub]

Mick, Turbo et al, I stand corrected re dates